Our laboratory has recently made an interesting series of observations that may have important implications for HIV vaccine design. Surprisingly, we have discovered that CD8+ T cells recognized Gag-derived epitopes by two hours post-infection: before integration, viral protein synthesis, and Nef-mediated MHC-I down-regulation. Additionally, CD8+ T cells recognized Tat and Nef-derived epitopes as early as 12-13 hrs post-infection. In contrast, epitopes derived from Env did not appear on the cell surface until eighteen hours post-infection, close to the onset of progeny virus production. We have also shown that Nef down-regulates MHC class I at 12-13 hrs post-infection. Thus, our preliminary data suggest that a CD8+ T cell-based HIV vaccine should stimulate responses against Gag, Tat and Nef. Conversely, Env-specific CD8+ T cells may not be effective because epitopes derived from this highly variable glycoprotein are expressed too late in the viral replication cycle well after Nef has down-regulated MHC-I. We hypothesize that vaccine induced CD8+ T cell responses directed against epitopes that are present on the cell surface soon after infection and prior to Nef-mediated MHC-I down regulation will control replication of the highly pathogenic SIVmac239 isolate. We will test this hypothesis in two related specific aims. Our Specific Aims (SA) are, therefore, to: I. Determine the exact timing after infection of when CD8+ T cell epitopes from each of the viral proteins are present on the surface of infected cells. II. We will use the information generated in SAI to design a DNA, Ad5 vaccine regimen. We will deliver proteins that encode epitopes present on the cell surface soon after infection as both whole ORFs and as minigenes. Delivery of these constructs as minigenes should broaden the immune response and facilitate control of SIVmac239 replication. We are very excited about our new findings in the rhesus macaque implicating Gag-specific CDS positive T cell responses in control of viral replication. While entire ORFs have been used in human vaccine modalities, dividing these larger ORFs into smaller pieces to avoid immunodominance has never been rigorously tested. The novelty of this proposal, therefore, lies in the intriguing, as yet unexplored possibility that subdominant epitopes in Gag and other viral proteins encoding epitopes expressed on the cell surface soon after infection could represent novel immunogens in our vaccine strategies for HIV. We feel that this proposal represents a highly significant and innovative approach to address the HIV epidemic.